print("Before:", encoded_1) encoder.update(state_1, state_2, reward_1, reward_2, wasserstein) encoded_1 = encoder.predict(state=state_1) print("After:", encoded_1) ################################################################################# ############################ Test Dynamics ###################################### state = np.random.rand(state_dim) next_state = np.random.rand(state_dim) action = np.random.rand(action_dim) encoded_state = encoder.predict(state=state) encoded_next_state = encoder.predict(state=next_state) next_state_from_dynamics = dynamics.predict(state=encoded_state, action=action) print("Before:", next_state_from_dynamics) dynamics.update(state=encoded_state, action=action, next_state=encoded_next_state) next_state_from_dynamics = dynamics.predict(state=encoded_state, action=action) print("After:", next_state_from_dynamics) #################################################################################